this is the best day of my life. something finally came along to trounce your purchase of that $4 louis-vitton-rip-off bratz baseball cap, eh? ASTOUNDING. =D
(btw, i have a picture of you and i in zeller's, wearing those hats, that you MUST see.)
If you were to take a topological approach and construct the set {y,yy,yyy,...} you could find a compact and totally bounded subset, {x1,x2,x3,...}. In which there exists a subsequence {i1,i2,i3,...} such that xin -> y. If we can prove that the real line is Hausdorff, then by the property of 2^n -> infinity, we can show that if xin converges to any y, it would have to be 2. Proof: given any two distinct points in the real line, say a & b where a < b, we can always find a number c that lies between a & b (because the real line is dense) and thus our two open sets, (a-2,c) and (c,b+2) are disjoint non-empty sets containing a & b respectively. Therefore, the real line is Hausdorff.
QED.
The other approach was to differentiate with respect to x, but I'm not good at differentiation.
Comments 17
something finally came along to trounce your purchase of that $4 louis-vitton-rip-off bratz baseball cap, eh? ASTOUNDING. =D
(btw, i have a picture of you and i in zeller's, wearing those hats, that you MUST see.)
Reply
QED.
The other approach was to differentiate with respect to x, but I'm not good at differentiation.
Reply
Wrong cageordy...
iwll be carefuul
Reply
1
Reply
1
Reply
Leave a comment